Loss Simulation Model Working PartyGoals & Progress Report
Mark R. Shapland, FCAS, ASA, MAAAConsulting Actuary
2007 Casualty Loss Reserve SeminarSan Diego, California
September 10-11, 2007
Sponsored by DRMC in 2005, began work in 2006.
Purpose: Create a simulation model that will generate claims that can be summarized into loss development triangles and complete rectangles.
Deliverables: Open source program available to CAS members, CAS seminar, and paper documenting work.
Time Frame: Test, refine and document prototype in 2007. Begin alternate versions in 2007, test & document in 2008.
Creation of LSMWP
Triangles by layer, claim information type (e.g., paid, incurred, Salv & Sub, claim counts, etc.), hazard, line of business, etc.
NOT focusing on testing of reserving methods, but creating simulated data sets for future research.
Primary criterion for judging quality of model is realism – i.e., data cannot be distinguished statistically from real data sets.
Procedures to review and test proposed modifications.
Goals of LSMWP
WP Co-Chairs:
Robert BearMark Shapland
Group A:Literature & Test Criteria
Curtis Parker
Group B:Data, Parameters & Testing
Joseph Marker
Group C:Model Development
Richard Vaughn
LSMWP Organization
Group A: Literature & Test Criteria (Curtis Parker)
– Survey existing literature and prepare bibliography.
– Develop testing criteria for using simulated data in reserve model testing.
• Necessary to assure model will support future research.
– Develop testing criteria for determining “realism” of simulated data.
• Can simulated data be statistically distinguished from actual data?
• Ultimate test: “DRM Challenge”
LSMWP Organization
Group B: Data, Parameters & Testing (Joe Marker)
– Identify data sources: Create spreadsheet of test data.
– Develop parameters for data sources.
– Test model and data using criteria from Group A. Group C: Model Development (Richard Vaughan)
– Evaluate modeling options and develop simulation model in at least two software environments. Open source for future enhancements.
– Refine and enhance model as a result of feedback from Group B. Document for CAS paper.
LSMWP Organization
Modeling individual transactions rather than aggregate triangles:– Don’t need triangle time intervals.– Can be used to test individual loss models.– Avoids criticism that “Of course that model predicts
that simulated data well, the simulated data is based on that model!”
– Easier to test against real data. Use intervals of one day for lags and waiting periods. Simulate each event usually in a claims system. Output in transaction details or a higher level of
aggregation, such as loss triangles.
Modeling Approach
Surveyed the literature on loss simulation and use of simulation to test reserving methods.
It is working to refine an approach to test the “realism” of simulated triangles.
The current approach is summarized in an upcoming ASTIN paper by Glenn Meyers titled “Thinking Outside the Triangle.”
Status Report – Group A
Working closely with a Data Source that has provided data for testing purposes.
Ball State University students have done the necessary database work and are busy estimating parameters for the prototype model.
Will complete testing the prototype and alternative versions after the Ball State course.
Will suggest model refinements and document test results.
Status Report – Group B
Developed and documented an APL prototype model. Expect the APL version will available to CAS members
with documentation in 2007. Work on a Visual Basic version is underway.
– The interface and model features will differ somewhat due in part to differences in software capabilities.
– This version will be fully tested and documented in 2008.
A version in the free statistical package R may also be developed. R is used by Group A for statistical testing and Group B for parameter estimates.
Status Report – Group C
Basic Model Underlying Prototype
Observation period: Assume that relevant loss process involves accidents or occurrences between an earliest accident date t0 and a latest accident date t1. The simulator tracks transactions from these accidents until settled.
Time intervals: Assume that parameters are constant throughout calendar months but may change from one month to next. Lags are measured in days.
Exposures: The user may specify a measure of exposure for each month. By default, the system assumes constant unit exposure. The purpose of the exposure parameter is to allow the user to account for a principal source of variation in monthly frequencies.
Basic Model Underlying Prototype
Events: Each claim may be described by the dates and amounts of the events it triggers: the accident date, the report date and an initial case reserve, zero or more subsequent valuation dates and case reserves changes, zero or one payment date and amount, and zero or one recovery date and amount.
Distributions: For most of variables, the user may specify a distribution and associated parameters. For convenience, the prototype model uses one or two parameter distributions with finite first and second moments and parameterizes them with their mean and standard deviation.
Frequency: Monthly claim frequency is assumed to have a Poisson distribution with mean proportional to earned exposure, or a Negative Binomial distribution with mean and variance proportional to earned exposure.– Accident dates for claims incurred in a month are distributed
uniformly across the days of that month.
Report lag: The lag between occurrence and reporting is assumed to be distributed Exponential, Lognormal, Weibull, or Multinomial. The Multinomial distribution allows the user to define proportions of claims reporting within one month, two months, and so on.
The lags between reporting and payment, between one valuation date and the next, and between payment and recovery or adjustment, are also assumed to be distributed Exponential, Lognormal, Weibull, or Multinomial.
Size of loss: The actual size of the loss to the insured, independent of responsibility for payment, is distributed Lognormal, Pareto, or Weibull.
Basic Model Underlying Prototype
Case reserve factor: Case reserves are assumed to equal the actual size of loss, adjusted for the minimum, the maximum, the deductible, and the probability of closure without payment, all multiplied by an adequacy factor. This factor is assumed to be distributed Lognormal, with mean and standard deviation specified by the user. The user may specify the mean at four separate points in time between the report and payment dates.
Fast-track reserve: User may specify a value to be assigned to each loss at first valuation, independent of regular case reserves & case reserve factor.
Initial payment factor: The initial payment of each loss not closed without payment is assumed to equal the actual size of loss, adjusted for the minimum, the maximum, the deductible, and whether or not the claim is closed without payment, multiplied by a payment adequacy factor (PAF). The PAF determines the size of any subsequent adjustment or recovery.
Basic Model Underlying Prototype
Second-level distributions: The LSMWP models the drift in parameter values that may take place for many reasons but chiefly because of business turnover. It has developed an autoregressive model to reflect parameter drift.
Monthly vectors of parameters: For nearly all distributional parameters, the user may specify a single value or a vector of values, one for each accident month or one for each development month, depending on the parameter involved.
Frequency Trend and Seasonality: The user may specify monthly trend and seasonality factors for frequency. These factors apply to the respective means in addition to all other adjustments.
Basic Model Underlying Prototype
Severity Trend: The user may specify monthly trend factors for severity.
• The “main” trend is allowed to operate up to the accident date and a fraction of this trend, defined by Butsic’s “alpha” parameter, is allowed to operate between accident and payment dates.
• Case reserves before the adequacy factor are centered around the severity trended to the payment date.
Basic Model Underlying Prototype
Lines and Loss Types: The prototype model recognizes that loss data often involves a mixture of coverages and/or loss types with quite different frequencies, lags, and severities. Therefore, it allows the user to specify a two-level nested hierarchy of simulation specifications, with one or more “Lines” each containing one or more “Types”.
• A typical Line might be “Auto,” typical Types within that Line might be “APD”, “AL-BI”, and “AL-PD.”
• This hierarchy allows the user to set up any reasonable one or two level classification scheme.
• Accident frequencies are modeled at the Line level and loss counts per accident are distributed among Types using a discrete distribution.
Basic Model Underlying Prototype
Lines and Loss Types Example: An Automobile occurrence might give rise to a single APD claim with probability 0.4, to a single AL-PD claim with probability 0.2, to a single APD and a single AL-PD claim with probability 0.2, to a single AL-BI claim with probability 0.1, to two AL-BI claims with probability 0.05, etc.
Correlations: The prototype model makes it possible to request correlated samples of certain variables without fully specifying their joint distribution. For the moment these variables are (a) the mean frequencies across Lines and (b) the size of loss and report lag within a Type.
Basic Model Underlying Prototype
Clustering: The prototype simulator allows a selectable fraction of loss sizes and a selectable fraction of case reserves to be rounded to two significant digits, imitating clustering around round numbers frequently observed.
Output: The prototype simulator produces output as tab-delimited text files or by launching an instance of Excel and populating it with worksheets. In both cases, the possible output tables include claim and transaction files (together displaying the complete loss history), all the usual triangles, a table of large losses, a summary of the simulation specifications, and a summary of the frequency derivation by month.
Basic Model Underlying Prototype
The LSMWP has made considerable progress in developing a model that we hope will become a valuable tool in researching reserving methods and models. Stay tuned!
We hope that actuaries will use this model to:– Better understand the underlying loss process.– Determine what methods and models work best in different
reserving situations.
Summary
